• Please review our updated Terms and Rules here

IBM System/23 keyboard reproduction

Just wanted to refresh this thread and see how things were going!

I'm anxious to start the restoration of the 3 tower units I have. Found some 15 pin IBM terminal monitors (3173, etc) in 5151 type shells in my storage building, gifts from my son, that I think will work provided I wire them up correctly and probably supply external power to. They appear to run off 12v/5v DC. I'll have to either inject 12v DC through a DB15 dongle or put a power jack on the monitor for it.
I've just ordered the keyboard PCBs (thanks to PCBWAY!), and I've also designed a simple DB25 breakout board for my logic analyser so I can see what the signals my Teensy is generating easily.
 
Keyboard PCBs have arrived. Need to order the Teensy 4.1's still (waiting for payday). And I need to figure out how to figure out the bus protocol.

IBM_Datamaster_Keyboard.jpg

I also designed some DB25 breakouts/passthroughs so I can easily hook my logic analyser up to the port to check the signalling ...
 
So, ahem, I made a small design mistake ... the Teensy 4.1 is 3.3V logic and I designed for 5V logic (using a 74HCT245 and 74HCT125) so I've not proceeded with this PCB build so far.
This probably isn't an issue Teensy -> IBM but IBM -> Teensy is a problem.

But looking at it again ... maybe I can still re-use the design & PCB.

[1] Swap the 74HCT125 for a 74LVC125
[2] Pull up rows to 3.3V instead of 5V

I've also got some NOS Intel 8048s on their way (arriving next week actually) so another idea was to just re-use the existing firmware & Intel 8048, although not sure what changes would be needed as the original keyboard was capacitive?
 
Checking in to see how things are going. Would love to have a working System 23 tower to show off at VCF Southwest at the end of May.
Hello,

It would be difficult at this stage, especially if no action has been taken until now. However, we could rush maintenance to see if at least one could be ready by then. About the keyboard, it would be very difficult too, but if what you want is to show the boot process, I can teach you how to trick the motherboard instead.

In any case, you have my support. Ask anything you need... but I think you have started the process too late for this deadline. Build the diagnostics probe first and try to find any fault the computer has. This has more priority than having a keyboard.

Regards
 
It is, I need to get around to building the test keyboard. I had to design a SOIC-14-to-DIP-14 converter PCB, which I now have fabricated. So ... I have no excuses now to not build and test one out (although I don't have an actual System/23 to test with, but we'll figure that out ...)
The interface is pretty simple:

Use a 8255 PPI in mode 1 input or emulate it.

Is this an ongoing project?
Yes! But since we have a lot on our plates it proceeds at a snail's speed, so...
Why do you ask? Do you have a Datamaster in need of a keyboard?
 
I am restoring 1 for the Large Scale Systems Museum. I am also writing a System/23 Datamaster emulator. My System/23 Datamaster emulator
I think we talked in private some time before.

You made another emulator, and with the FDC? HATS OFF! I am no longer the most knowledgeable user here... not that I care, I am very happy about it.
Is it stand-alone or is attached to some multi-emulator like mine?

I am a little sad because you could have done it faster if you asked me, since I know many of the details to make the tests pass. That's especially true with 05h or the RAM tests.

In any case, I am very happy because the System/23 is being worked on. We can't provide a replacement keyboard yet, but I can teach you how to trick the motherboard to think it has one. What do you think?

Oh, I would also like to make you a request, if possible. Please, would you share the details of the firmware of the machines you are working with? In case we lack a specific piece of firmware, could you shate the dump with us, please? I ask because before me there was only a set of ROMs dumped, had a defective one; then I arrived, dumped everything I could identify, then I started to ask other users and owners for their dumps... and that got to the current firmware backup file we have today. Since the Datamaster ROMs get bad very easily, it is a lifesaver for other users and a future warranty for yourself. Consider it also some kind of initiation rite.

Thank you for the good news!
 
Mine is stand alone. The Rom install on the System/23 that I am restoring is 1.04. I was given rom dumps from a guy in Europe that wrote the firmware for the System/23. I have dumps for 1.01, 1.03, 1.04 and 1.05. I am looking for more floppy disk images.
 
Mine is stand alone. The Rom install on the System/23 that I am restoring is 1.04. I was given rom dumps from a guy in Europe that wrote the firmware for the System/23. I have dumps for 1.01, 1.03, 1.04 and 1.05. I am looking for more floppy disk images.
I think those are our dumps. How do I know? Because we dumped everything with the exception of 1.02, that it has still to appear.
 
Oh, it doesn't look like 74LVC125 is available as a DIP-14 so I've whipped up a SOIC-14 to DIP-14 adaptor board so I can re-use the existing keyboard PCB.

View attachment 1320544

You can take exactly the same approach for a microcontroller if you decide to try making a non-Teensy controller at some point. If your PCB design is double sided, you can sneak supporting passive components (SMD/SMT) onto the other side to save space.

If you really wanted to, you could even route all the connections to a single line of pins and mount the uC carrier to the mainboard without a DIP style socket/header.

-----

If finding a DB-25 connector became difficult, you you could always use a keyed IDC connector. They are sufficiently snug, secure for this kind of usage.
 
Last edited:
Back
Top